Nonlinear Regression Modeling to Predict Thermal Endurance of XLPE Material under Thermal Aging

Y. Kemari , A. Mekhaldi (1) , M. Teguar (1) , G. Teyssedre (2)

1 LRE - Laboratoire de recherche en électrotechnique = Electrical engineering research laboratory [ENP, Alger]
2 LAPLACE-DSF - Diélectriques Solides et Fiabilité

In this work, the mechanical behavior of XLPE under accelerated thermal aging is investigated. The tensile strength and elongation at break are used as aging markers in order to assess the degradation level of the insulation. In the aim to describe the evolution of the mechanical properties with aging and to estimate the insulation lifetime, a nonlinear regression model is proposed. The model parameters have been estimated using the method of least squares. The experimental results show that thermal aging leads to mechanical degradation of XLPE by decreasing its tensile strength and elongation at break. The critical deterioration of the material was reached at 120 and 140 °C. The validity of the regression model was verified and the results show that one-phase exponential decay expression could well fit the experimental data. In addition, thermal endurance of XLPE was investigated using the Arrhenius law. On the other hand, permittivity and AC conductivity have been monitored in order to study their correlation with thermal aging.

Mots clés: Electrical insulation; dielectrics; thermal aging; cross-linked polyethylene; power cables; nonlinear regression; thermal endurance. I

Domaines: Energie électrique

Y. Kemari, A. Mekhaldi, M. Teguar, G. Teyssedre. Nonlinear Regression Modeling to Predict Thermal Endurance of XLPE Material under Thermal Aging. 2018 IEEE 2nd International Conference on Dielectrics (ICD), Jul 2018, Budapest, Hungary. pp.1-4, ⟨10.1109/ICD.2018.8514644⟩. ⟨hal-02396153⟩